Prodrugs of neuron-selective monoamine oxidase inhibitors: Amino acid derivatives of 1-(4-aminophenyl)-2-aminopropanes

Article abstract of DOI:10.1016/S0223-5234(99)80047-6

Six amino acid derivatives of 1-(4-aminophenyl)-2-aminopropanes and their parent amines were synthezised and tested for their potency and selectivity in inhibiting monoamine oxidase (MAO) in vitro and in vivo. The amino acid derivatives were 300-1000 times less potent than the parent amines in inhibiting the MAO-A activity in a rat brain mitochondrial preparation in vitro. All compounds, except the (R)-valinamide derivative (22), were potent inhibitors of MAO in the rat brain in vivo and were, like the parent amines markedly more potent within the monoaminergic neurons than in other neurons. The glycinamide derivative 7 showed the largest difference between intra- and extra-neuronal inhibition in serotonergic neurons. The time course of the inhibitory effect of 7 in vivo showed that it is a reversible inhibitor with a long duration.

Full text of DOI:10.1016/S0223-5234(99)80047-6

Eur. J. Med. Chem. 34 (1999) 137−151
© Elsevier, Paris
137
Original article
Prodrugs of neuron-selective monoamine oxidase inhibitors:
amino acid derivatives of 1-(4-aminophenyl)-2-aminopropanes
Lennart Florvalla, Ingrid Fagervallb, Lars-Gunnar Larssonb, Svante B. Rossb*
aDepartment of Medicinal Chemistry, Preclinical R&D, Astra Arcus AB, 15185 Södertälje, Sweden
bDepartment of Biochemical Pharmacology, Preclinical R&D, Astra Arcus AB, 15185 Södertälje, Sweden
(Received 19 March 1998; accepted 14 October 1998)
Abstract – Six amino acid derivatives of 1-(4-aminophenyl)-2-aminopropanes and their parent amines were synthezised and tested for their
potency and selectivity in inhibiting monoamine oxidase (MAO) in vitro and in vivo. The amino acid derivatives were 300–1000 times less
potent than the parent amines in inhibiting the MAO-A activity in a rat brain mitochondrial preparation in vitro. All compounds, except the
(R)-valinamide derivative (22), were potent inhibitors of MAO in the rat brain in vivo and were, like the parent amines markedly more potent
within the monoaminergic neurons than in other neurons. The glycinamide derivative 7 showed the largest difference between intra- and
extra-neuronal inhibition in serotonergic neurons. The time course of the inhibitory effect of 7 in vivo showed that it is a reversible inhibitor
with a long duration. © Elsevier, Paris
monoamine oxidase inhibitors / amiflamine / prodrug / serotonin / noradrenaline / dopamine
1. Introduction
The enzyme monoamine oxidase (MAO; amine: oxy-
gen oxidoreductase (deaminating, flavin-containing); EC
1.4.3.4) exists in two forms, A and B, with different
substrate specifications. 5-Hydroxytryptamine (5-HT) is
a
prefered substrate for the
A
form and
â-phenylethylamine (PEA) for the B form. Since there
are two forms of the enzyme, it is possible to develop
compounds that selectively inhibit the A form and leave
the B form active to deaminate amines, like tyramine,
ingested in the food. Although the B form is abundant in
the brain and the A form in peripheral tissues, the A form
is functionally responsible for the main part of the
deamination of 5-HT, noradrenaline (NA) and dopamine
(DA) in the aminergic neurons in the brain [1, 2]. During
the last two decades a new generation of selective and
reversible MAO-A inhibitors has been developed among
which moclobemide is clinically used as an antidepres-
sant drug (for a review see [3]).
Amiflamine (1a) and related p-amino substituted am-
phetamine derivatives have been shown to preferably
inhibit the MAO activity within the monoaminergic
neurons by being accumulated within these neurons via
the membranal amine transporters [4–7]. Because of their
high concentration within these neurons, these com-
pounds inhibit MAO therein at considerably lower doses
than those inhibiting MAO in cells lacking these trans-
porters [7]. The purpose of the present study was to
*Correspondence and reprints

138
Table I. 1-(4-Amino-2-fluorophenyl)-2-aminopropane and amino acid derivatives.
Compound
R1
R2
R3
R4
m.p., °C
Formula
2
3
4
5
6
7
8
21
22
H
H
H
F
F
H
F
H
H
H
Me
Me
H
H
H
H
H
H
Me
Me
Me
Me
H
Mé
Me
H
Me
Me
Me
Me
208–209
> 180 (D)
148–149
106–107
102–103
112–113
183–184
246–248
225–226
C10H14F2N2 × 0.5(COOH)2
C9H12F2N2 × 1.5(COOH)2 × H2O
C10H15FN2 × HAc
C12H17F2 N3 O × HAc
C11H16FN3 O × HAc
C12H15FN3O × HAc
C13H20FN3O × (COOH)2
C17H29N3O × 2HCl
Glycyl
Glycyl
Glycyl
Glycyl
(S)-Valyl
(R)-Valyl
C17H29N3O × 2HBr
a
b
Configuration (R,S); configuration (S).
synthesize low-potency prodrugs to neuronselective
MAO-A inhibitors structurally related to amiflamine. The
advantage of this approach would be that the interaction
between ingested tyramine and MAO inhibitor in the
intestines should be minimized. It is likely that the
MAO-A activity in the intestinal mucosa is responsible
for the main part of the deamination of tyramine ingested.
Consequently, a useful pro-drug should not be deassoci-
ated in the intestines but be taken up into blood circula-
tion, forming the active compound in the liver or the
brain. In previous works with amiflamine analogues, we
found that 2-fluoro substitution, e.g. 1b and 1c, gave very
high MAO-A inhibitory potency in vitro and in vivo [4, 7,
8]. Hence these structures were chosen as the basis for the
synthesis of a series of glycinamides as potential pro-
drugs. In this report we describe the synthesis and
pharmacological testing of these compounds as MAO-A
inhibitors with neuron selective action in vivo. In addi-
tion, the S- and R-valine amides of amiflamine ((S)-1a)
were prepared and evaluated as MAO-inhibitors.
difluorobenzoyl chloride by treatment with thionyl chlo-
ride and then reacting with concentrated aqueous ammo-
nia to give the amide. Degradation of the obtained amide
via the Hofmann reaction gave the corresponding aniline.
The desired intermediate 9 was then obtained by the
reaction of the aniline with formic acid. Compound 9 was
2. Chemistry
The target fluoro compounds prepared in this study are
listed in table I and their synthesis are outlined in
figures 1–6.
Compound 2 was prepared as illustrated in figure 1.
3,5-Difluorobenzoic acid was converted to 3,5-
Figure 1. Reagents: (a) SOCl2; (b) NH3; (c) Br2/NaOH; (d)
HCOOH; (e) LAH; (f) t-BuCOCl; (g) n-BuLi; (h) DMF; (i)
MeCH2NO2.

139
Figure 2. Reagents: (a) t-BuCOCl; (b) n-BuLi; (c) DMF; (d) BnCl/K2CO3; (e) MeCH2NO2; (f) LAH; (g) H2, Pd/C.
also prepared from commercially available 3,5-
difluoroaniline and formic acid. The formanilide 9 was
reduced with lithium aluminium hydride and the obtained
N-methylaniline was used directly in the next step to
yield the N-pivaloylamide 10. The pivaloyl protecting
group was considered to be ideal since the absence of
acidic α-hydrogens excludes undesired lithiations of the
amide alkyl residue [9]. Furthermore the pivaloyl group
proved sufficiently labile for removal in the final lithium
aluminum hydride reduction step.
Lithiation of compound 10 with butyllithium and then
reaction with N,N-dimethylformamide gave the desired
aldehyde 11. In the next step the 2-nitropropene interme-
diate was prepared by the condensation of compound 11
with nitroethane in the presence of ammonium acetate.
Preparation of the target amine 2 was then achieved by
the treatment with lithium aluminum hydride, which in
one step effected both reduction and deprotection of the
2-nitropropene intermediate. Keeping the pivaloyl amide
in excess and performing the reaction at low temperature
was found to be important for the outcome of the
deprotection.
benzylated directly to the corresponding N-benzyl-N-
methyl derivative. The intermediate aniline was not
isolated at this stage, but formylated by the Vilsmeier–
Haack procedure to the desired intermediate 4-(N-benzyl-
N-methylamino)-2-fluorobenzaldehyde. The condensa-
tion of this compound and nitroethane in the presence of
ammonium acetate gave the 2-nitropropene derivative 18.
Reduction of 18 with lithium aluminium hydride and
debenzylation by catalytic hydrogenation of the obtained
amine, afforded compound 4. Reduction of 18 with
lithium aluminum hydride and chloracetylation of the
obtained amine with chloroacetyl chloride, gave amide
19. Amination of 19 by means of dibenzylamine yielded
In an attempt to synthesise compound 3, lithiation and
DMF treatment of 3,5-difluoro-N-pivaloylaniline was
found to give rise to the benzaldehyde 14 in figure 2.
Using a different strategy 3,5-difluoroaniline was N,N-
dibenzylated with benzyl chloride in the presence DMF
and anhydrous potassium carbonate. The N,N-
dibenzylaniline 15 was reacted with butyllithium and
DMF to give the desired aldehyde 16. The 2-nitropropene
derivative 17 was then prepared by the condensation of
compound 16 and nitroethane in the presence of ammo-
nium acetate. Selective reduction of the 2-nitropropene
group with lithium aluminum hydride and subsequent
debenzylation in situ by catalytic hydrogenation in the
presence of palladium, gave the desired target amine 3.
The synthesis of compound 4 in figure 3 started from
3-fluoroaniline, which was reacted with formic acid. The
formed formanilide was reduced with lithium aluminum
hydride and the obtained 3-fluoro-N-methylaniline was
Figure 3. Reagents: (a) HCOOH; (b) LAH; (c) BnCl; (d)
POCl3/DMF; (e) MeCH2NO2; (f) H2, Pd/C; (g) ClCH2COCl;
(h) HNBn2.

140
Figure 5. Reagents: (a) ClCH2COCl; (b) HNBn2; (c) H2, Pd/C;
(d) NH3.
compound 21 was obtained by catalytic hydrogenation.
Starting from (R)-N-carbobenzyloxyvaline [12] com-
pound 22 was prepared in a similar procedure. However,
in this case the deprotection was effected by using
hydrobromic acid (figure 6).
Figure 4. Reagents: (a) LAH; (b) BnCl; (c) n-BuLi; (d) DMF;
(e) MeCH2NO2; (f) ClCH2COCl; (g) HNBn2; (h) H2, Pd/C.
3. Pharmacology
3.1. MAO inhibition in vitro
amide 20, which was debenzylated by catalytic hydroge-
nation to give the target glycinamide derivative 7.
The preparation of compound 5 in figure 4 started from
the formanilide 9 which was reduced with lithium alumi-
num hydride, benzylated and treated with butyllithium
and N,N-dimethylformamide to give the desired aldehyde
12. The conversion of compound 12 into the glycinamide
5 was then effected by the same route used for the
preparation of compound 7. It may be noted that an
attempt to transform the intermediate N-benzyl-N-
methyl-3,5-difluoroaniline into aldehyde 12 by means of
the Vilsmeier–Haack formylation process, was unsuc-
cessful.
The results are given in table II. They show that the
fluorinated aminopropanes (1b, 1c, 2, 3, 4) were selective
and potent inhibitors of the A form of MAO in vitro, since
they inhibited the deamination of 5-HT at much lower
concentrations than those required to inhibit the deami-
nation of PEA. The di-fluoro derivative 3 was 10 times
more potent than the corresponding mono-fluoro com-
The preparation of compound 8 started from 1-(4-dime-
thylamino-2-fluorophenyl)-2-aminopropane [4] which
was treated with chloroacetyl chloride (figure 5). Amina-
tion of the obtained amide with ammonium hydroxide at
room temperature gave the target glycinamide. Chloro-
acetylationof1-(4-dibenzylamino-2-fluorophenyl)-2-ami-
nopropane [4], amination of the yielded chloroacetamide
with dibenzylamine and subsequent catalytic hydrogena-
tion, gave the desired compound 6 (figure 5).
The preparation of compound 21 was effected by the
acylation of (S)-(+)-1-(4-dimethylamino-2-methylphe-
nyl)-2-aminopropane [10] with (S)-N-carbobenzyloxy-
valine [11] in the presence of CDI. The deprotection of
the obtained carbobenzyloxy intermediate to the target
Figure 6. Reagents: (a) CDI, (S)-(+)-1-(4-dimethylamino-2-
methylphenyl)-2-aminopropane; (b) H2, Pd/C; (c) HBr.

141
Table II. Inhibition of MAO in vitro and inside (IN) and outside (EN) the monoaminergic neurons ex vivo.
a
Compound
In vitro inhibition
Ex vivo inhibition
ED50 (µmol/kg p.o.)
5-HT
IC50 (µM)
PEA
5-HT
NA
IN
DA
IN
NA
IN
EN
EN
EN
b
1b
800
0.75
>100
0.08
27
5.2*
14*
1.1*
0.7*
0.45*
0.6*
0.5*
0.5*
1.2*
48
>70
5.8
12
3.0
2.3
3.8
2.0
4.4
0.8*
1.8*
1.2*
1.2*
0.45*
0.7*
0.24*
0.2*
0.5*
nt
57
0.8*
nt
1.0*
1.0*
1.1*
nt
nt
nt
nt
nt
4.4
nt
6.2
6.3
2.2
nt
nt
nt
nt
nt
6
4
7
1c
8
3
2
5
21
>100
>100
>100
120
>100
>100
40
>100
>100
>100
>1000
>70
5.4
5.0
3.0
3.5
6.2
2.9
2.7
nt
c
1.2
>100
0.07
0.01
4.5
>100
>100
0.8
1.7*
>>17 (2 %)
1.1*
17
22
>>17 (2 %)
5.2
nt
2.0*
nt
8.0
nt
3.3*
nt
4.4
Amiflamine (1a)b
a Test compounds administered 1 h before the subcutaneous injection of phenelzine and the rats were sacrificed 2 days later. b Data from [7].
c
Data from [4] and [5]. *: denotes significant (p < 0.05) larger inhibition inside than outside the neurons (Mann–Whitney U-test); nt = not
tested. Between brackets: % inhibition at the dose noted.
pound 1b. The 4-methylamino derivative 4 had 10–15
times higher potency than 1b. The 4-methylamino-2,6-
difluoro derivative 2 had the highest potency to inhibit
both forms of the enzyme. The glycinamide derivatives
(5–8) were 300–1000 times weaker than the correspond-
ing amines (table II). The inhibition curves of 4 and 7 are
shown in figure 7. The (S)- and (R)-valine derivatives of
amiflamine (21, 22) also had weak MAO inhibitory
potencies in vitro compared with amiflamine (1a) itself
(table II).
after 1 mg/kg (3.3 µmol/kg) p.o. are shown in figure 9.
The inhibition in serotonergic neurons was significant 8 h
but not 20 h after the administration. In noradrenergic
neurons the inhibition was still significant (about 40%)
3.2. MAO inhibition in vivo
All compounds except the (R)-valine derivative of
amiflamine 22 were potent inhibitors of MAO in vivo
(table II). They were all more potent in the monoamin-
ergic neurons than in other neurons or cells, i.e. had
neuron-selective action. With the exception of 5 and the
primary 4-amino derivatives Ib the compounds were
about equal active in the three types of aminergic neuron
systems examined. These primary 4-amino derivatives
had considerably lower potency in the 5-HT neurons than
in NA and DA neurons but had quite high intraneuronal
selectivity in the two former systems. The glycinamide
derivative 7, related to the 4-methylamino-2-fluoro com-
pound 4, had the highest selectivity with regards to the
inhibition of neuronal 5-HT deamination (table II, fig-
ure 8). This compound was therefore chosen for further
studies. The time courses of the MAO-A inhibition in
serotonergic and noradrenergic neurons in hypothalamus
Figure 7. Inhibition by 7 (squares) and 4 (circles) of the
deamination of 14C-5-HT (MAO-A) (solid symbols) and 14C-
phenethylamine (MAO-B) (open symbols) in vitro. A mito-
chondrial preparation from rat brain was used as enzyme. The
concentration of 14C-5-HT was 50 µM and that of 14C-PEA was
2.5 µM.

142
Table III. The effect of compound 7 on the monoamine levels in hypothalamus, hippocampus and striatum 1 h after oral administration.
Dose (µmol/kg)
Amine concentration, per cent of controls, ± S.E.M.
Hypothalamus
Hippocampus
Striatum
5-HT
5-HT
NA
DA
5-HT
NA
DA
c
a
a
b
a
b
3.3
16
127 ± 5
117 ± 3
116 ± 3
127 ± 7
122 ± 12
130 ± 8
159 ± 22
154 ± 12
129 ± 7
105 ± 4
122 ± 5
b
a
b
b
b
149 ± 11
146 ± 14
153 ± 11
Control values (nmol/g tissue): hypothalamus: 5-HT: 3.16 ± 0.05; NA: 9.21 ± 0.45; DA: 2.14 ± 0.29; hippocampus: 5-HT: 1.99 ± 0.06; NA:
1.71 ± 0.06; striatum: 5-HT: 1.57 ± 0.07; DA: 45.3 ± 1.8. a p < 0.05; b p < 0.01; c p < 0.001 (Dunnett’s t-test following ANOVA) compared
with saline controls.
after 20 h. Maximal effect was obtained 2 h after the
administration.
1 h after administration (data not shown). The duration of
the elevation of the NA concentration in hypothalamus
and hippocampus was somewhat longer compared with
that of the 5-HT concentration.
3.3. Effect on monoamines and their metabolites
In the first experiment the rats were killed 1 h after the
oral administration of 7 (3.3 and 16 µmol/kg). The 5-HT
level in the three regions analyzed (hypothalamus, hippo-
campus and striatum) increased with 25–30% of the
control values at the lower dose and with 45–50% at the
higher dose (table III). The NA level increased much
more in hippocampus (58%) than in hypothalamus (17%)
and both doses produced similar elevation. The DA level
in striatum was significantly increased only after the high
dose. The concentration of 5-HIAA hippocampus and
hypothalamus decreased with 20–25% and 45–55%,
respectively (table IV). DOPAC was strongly decreased
in hypothalamus (35% at 3.3 µmol/kg and 70% at
16 µmol/kg) and striatum (70 and 85%, respectively)
(table IV). The HVA levels were also markedly decreased
in these regions.
4. Discussion
In previous studies it has been shown that it is possible
to develop neuron-selective MAO-A inhibitors by utiliz-
ing the specific transport mechanisms in the the cell
membranes of the monoaminergic neurons [5–7]. By
being transported into the neurons, compounds like
α-ethyltryptamine, p-methoxyamphetamine, amiflamine
and related p-aminoamphetamine derivatives are accu-
mulated in the amine neurons and thereby are inhibiting
MAO-A therein at lower doses than those inhibiting
MAO-A in cells or neurons lacking these transporters [7].
Since the transporters are not identical in the various
amine neurons, the structural requirements for uptake is
different which makes it possible to obtain MAO-A
selectivity in a certain monoamine system, e.g. norad-
renaline neurons [7].
The time courses of the changes in the concentrations
of the monoamines and their metabolites after
3.3 mmol/kg of 7 were determined in a second experi-
ment. The maximal increase in the amine levels were
obtained 2 h after the administration whereas the concen-
trations of the metabolites had the lowest values already
The aim of the present study was to develop prodrugs
to amiflamine-like compounds with low in vitro affinity
for MAO-A. This was achieved by making amino acid
derivatives which had a hundred-fold lower in vitro
Table IV. Effect of compound 7 on monoamine metabolites in the rat hypothalamus, hippocampus and striatum.
Dose (µmol/kg)
Concentration of monoamine metabolites in per cent of controls, ± S.E.M.
Hypothalamus
5-HIAA
Hippocampus
Striatum
DOPAC
DOPAC
HVA
5-HIAA
5-HIAA
HVA
c
a
a
b
c
c
c
3.3
16
77 ± 3
45 ± 5
66 ± 3
31 ± 2
69 ± 5
35 ± 6
80 ± 5
49 ± 5
78 ± 4
55 ± 6
41 ± 4
15 ± 1
57 ± 4
28 ± 6
c
c
c
c
b
c
The concentration of 5-hydroxyindoleacetic acid (5-HIAA), 3,4-dihydroxyphenylacetic acid (DOPAC) and homovanillic acid (HVA) in brain
1 h after the oral administration of 6 were analysed as described in the experimental part. Control values (nmol/g tissue): hypothalamus:
5-HIAA: 2.54 ± 0.08; DOPAC: 0.44 ± 0.04; HVA: 0.25 ± 0.04; hippocampus: 5-HIAA: 1.98 ± 0.09; striatum: 5-HIAA: 2.31 ± 0.08; DOPAC:
a
b
c
6.15 ± 0.34; HVA: 4.67 ± 0.29. p < 0.05; p < 0.01; p < 0.001 (Dunnett’s t-test following ANOVA) compared with saline controls.

143
Figure 8. Dose–response curves of the in vivo inhibition by 7
of MAO-A inside (circles) and outside (squares) 5-HT neurons
(top) and NA neurons (middle) in rat hypothalamus and
DA-neurons (bottom) in rat striatum. Different doses of 7 were
given orally 1 h before the subcutaneous injection of phenelzine
sulfate (4 mg/kg in the experiment with 5-HT and NA and
2.5 mg/kg in the experiment with DA) and the rats were killed
48 h later. Hypothalami and striata were dissected out and
homogenized. The homogenate was incubated with 0.1 µM
Figure 9. Time course of the inhibition of MAO-A by com-
pound 7 within (circles) and outside (squares) 5-HT (top) and
NA (bottom) neurons in the rat hypothalamus. 7 (3.3 µmol/kg)
was administered orally at different times before the subcuta-
neous injection of phenelzine sulfate, 4 mg/kg, and the rats
were killed 48 h after the latter injection. The MAO activity
was determined and the inhibition at the times for the phenel-
zine injection was calculated as described in the experimental
section. Each value is the mean ± S.E.M. (bars) from 4 rats. *:
significant protection against phenelzine (p < 0.05; Mann–
Whitney U-test).
3
14C-5-HT, 0.25 µM H-NA or 0.25 µM 14C-DA in the absence
and the presence of selective uptake inhibitors. The protection
against the irreversible inhibition caused by phenelzine was
taken as a measure of the MAO inhibition that was calculated
as described in the experimental section. Each value is the mean
from 4 rats. *: p < 0.05 vs. phenelzine controls (Mann–Whitney
U-test).

144
potency to inhibit MAO-A than the parent drug. Since
these amides can be enzymatically hydrolyzed in vivo
forming the active inhibitors, it might be possible to
overcome the main obstacle with MAO-A inhibitors:
potentiation of tyramine-induced hypertension. This is
thought to involve MAO-A activity in the intestine
mucosa which is the first metabolic barrier for ingested
tyramine. Thus, a pro-drug of MAO-A inhibitor which is
able to pass into the blood circulation intact and thereafter
form the active compound in the liver may be a thera-
peutically useful antidepressant.
In accordance with previous observations [5–8], it was
found that the primary aminophenyl derivative 1b was
more potent in inhibiting MAO in noradrenergic than in
serotonergic neurons which indicates that the 1-(4-
aminophenyl)-2-aminopropanes are better solutes for the
noradrenaline transporter than for the serotonin trans-
porter whereas the mono- and dimethylaminophenyl
derivatives, e.g. 4 and 1c appear to be similar good
substrates for both transport systems. This structure–ac-
tivity relationship seems, however, to be revoked by
substituting the phenyl with two fluorine atoms at ortho
positions, e.g. compound 3 which had similar potency in
5-HT and NA neurons and the very potent compound 2
which tended to be most active in NA neurons.
The results obtained in this study show that the amino
acid derivatives of 1-(4-aminophenyl)-2-aminopropanes
were very weak MAO inhibitors in vitro but quite strong
and neuron-selective inhibitors in vivo. The lack of in
vivo activity of the R-valine derivative of amiflamine (22)
shows that enzymatic hydrolysis of the amide is neces-
sary for MAO inhibitory effect. The hydrolysis of the
amides may occur in the intestines by peptidases, in the
liver by amidases or at both sites. Although the present
study does not give a definite answer to this, the time
course of the in vivo inhibitory action of 7, the compound
chosen for extended study, with optimal effect 2 h after
oral administration indicated that the glycinamide was
mainly hydrolyzed in the liver. In accordance with this
view, it has been reported the midodrine (the glycinamide
of 2,5-dimethoxy-â-phenylethanolamine) is almost com-
pletely absorbed (90% bioavailability) after oral admin-
istration in man and is rapidly hydrolyzed in the liver
forming the parent amine [14]. Further studies have to be
performed in order to show if this pro-drug principle
might result in MAO inhibitors with potential clinical
value.
The MAO inhibition in vivo by the compounds was
established with two methods. The ex vivo deamination
of the labelled monoamines in synaptosomal preparations
from rats treated 2 days earlier with the test compound
before the injection of phenelzine measures the protection
by the test compound against the irreversible inhibition of
MAO by phenelzine. This protection is consequently a
measure of the inhibitory potency of the test compound
competing with phenelzine for the enzyme [13]. If the
enzyme activity is measured with low concentrations of
the labelled transmitter amine in the absence and the
presence of selective upptake inhibitors, it is possible to
measure the activity within a particular aminergic neuron
system [7]. This technique showed that the compounds
inhibited the MAO activity in monoamine neurons at
considerably lower doses than in other neurons or cells.
Thus, compound 7 was 17 times more potent in inhibiting
MAO-A in 5-HT neurons in hypothalamus than within
other hypothalamic cells. This preference for inhibition of
the aminergic MAO is obviously due to uptake of the
inhibitor by the amine transporters, resulting in high
concentration inside these neurons [6]. The increase in
the concentrations of the transmitter amines and the
decrease in their metabolites in various brain regions
confirm the MAO inhibitory action of 7 in vivo. The time
courses of the elevation of the amine levels followed
those of the inhibition within the neurons whereas the
decrease in the metabolites decayed more rapidly, prob-
ably due to extraneuronal deamination of the transmitter
amines.
5. Experimental protocols
5.1. Chemistry
Melting points were determined with a Mettler FP62
automatic melting point recorder. The elemental analyses
were performed by Mikro Kemi AB, Uppsala, Sweden.
Where analyses are indicated by symbols of the elements,
analytical results obtained for those elements were within
±0.4% of the theoretical values. Most of the intermediates
were analyzed by means of GC/MS and used directly in
the subsequent reaction step without any purification. Gas
chromatography was carried out on a HP 5890 instrument
It must be noted that the glycine derivatives studied
may not be the optimal compounds, since they were
synthesized from the racemic amines. The enantiomers
may differ in the rate of being formed by the enzymatic
hydrolysis of the prodrugs and also in the capability of
being actively transported into the monoaminergic neu-
rons. In addition, it has been shown that the (S)-
enantiomer of 1a (amiflamine) is about 4 times more
potent than the corresponding (R)-enantiomer as an
inhibitor of MAO-A in vitro [5].
1
with a 10 m × 0.1 mm CP-Sil-5 capillary column. H
NMR spectra were recorded with a Varian Gemini-300

145
NMR spectrometer. Mass spectra were obtained at 70 eV
on an ITD800 (C I mode, CH4) or a LKB291 (EI mode)
mass spectrometer.
solution of n-BuLi in hexane. After 0.5 h at –80 °C
4.5 mL (0.05 mol) of dry DMF in 10 mL of THF was
added dropwise at the same temperature and the mixture
was stirred for 1 h. The mixture was poured into dilute
H2SO4 and ice and the aldehyde was extracted with ether.
The extract was washed with water and the ether was
evaporated. The obtained crude aldehyde was purified by
means of the hydrogen sulfite addition compound as
follows. To the residue was added 5.0 g of NaHSO3 and
75 mL, of water. The mixture was stirred overnight at
room temperature and extracted with ether. The water
layer was separated and 10 g of K2CO3 was added. After
stirring for 0.5 h the mixture was extracted with ether.
The extract was washed with water, dried and the ether
evaporated to yield 2.7 g (51%) of an oil, containing 96%
(GC) of the wanted aldehyde. The oil solified when
stirred with cool petroleum ether. The obtained solid
aldehyde melted at 32–33 °C. MS (CI): m/z (M+ + 1) 256;
1H NMR (CDCl3) δ 10.3 (s, 1H, CHO), 6.9 (d, 2H, JH,F
= 9.3 Hz, Ar), 3.3 (s, 3H, CH3), 1.2 (s, 9H, 3CH3).
5.1.1. 3,5-Difluoroformanilide 9
Method I: A solution of 5.0 g (0.039 mol) of 3,5-
difluoroaniline in 15 mL of 98% formic acid was refluxed
for 2 h and then poured into 250 mL of crushed ice. After
stirring for 1 h the obtained precipitate was filtered,
washed with water and dried: Yield 5.0 g (82%); m.p.
88–89 °C. GC = 99%. MS (CI): m/z (M+ + 1) 158.
Method II: To a stirred mixture of 31.6 g (0.2 mol) of
3,5-difluorobenzoic acid, 200 mL of CH2Cl2 and a cata-
lytical amount of DMF SOCl2 (80 mL, 1 mol) was added
portionwise. The mixture was refluxed for 4 h, the solvent
was evaporated and the residual oil dissolved in 80 mL of
dry THF. The solution was added dropwise under stirring
and cooling in ice to 400 mL of concentrated NH4OH.
The obtained amide was collected by filtration and dried.
Yield 27.7 g (88%); m.p. 156–157 °C. The product
(27.7 g, 0.17 mol) was added in portions, whith stirring,
to an ice-cold solution of 10.6 mL Br2 and 33.3 g of
NaOH in 300 mL of water. The mixture was heated at
reflux temperature for 2 h and then extracted with ether.
The extract was dried with Na2SO4 and the ether was
evaporated. To the residual oil was added 70 mL of 98%
formic acid and the solution was refluxed for 2 h and then
worked up as described in Method I: yield 18.9 g (71%);
m.p. 88–89 °C; MS (CI): m/z (M+ + 1) 158. Anal.
(C7H5F2NO) H, N; C: calc. 53.51; found 53.0, 52.9.
5.1.4. 1-(2,6-Difluoro-4-methylaminophenyl)-2-amino-
propane semioxalate 2
A mixture of 10.5 g (0.04 mol) of compound 11,
5.3 mL (0.07 mol) of nitroethane and 5 g of ammonium
acetate in 100 mL of acetic acid was heated under reflux
for 1 h. The acetic acid was evaporated and after the
addition of water the mixture was extracted with ether.
The extract was washed with a NaHCO3 solution and
with water and dried over Na2SO4. The ether extract gave
when evaporated, 12.3 g of an oil. Examination of the
above residue by GC/MS indicated that the product
contained 70% of the wanted â-nitrostyrene. MS (CI):
m/z (M+ + 1) 313.
5.1.2. 3,5-Difluoro-N-methyl-N-pivaloylaniline 10
A solution of 5.0 g (0.03 mol) of compound 9 in 50 mL
of ether was added dropwise under nitrogen to a stirred
mixture of 3.0 g LAH in 50 mL of ether. The mixture was
stirred and heated under reflux for 3.5 h. After dropwise
addition of 15 mL of a saturated Na2SO4 solution while
stirring and cooling in ice-water, the mixture was filtered.
The filtrate was dried over anhydrous Na2SO4 and
evaporated. The residual oil was dissolved in 20 mL of
dry pyridine and 4.5 mL (0.04 mol) of pivaloyl chloride
was added in one portion while stirring. After 12 h in
room temperature the mixture was poured in 500 mL of
ice-water. The obtained precipitate was filtered, washed
with water and dried to give 9 (6.5 g, 96%); m.p.
69–70 °C. MS (CI): m/z (M+ + 1) 228.. Anal.
(C12H15F2NO) H, N; C: calc. 63.42; found 62.9, 62.6.
To a stirred solution of 12.3 g of the obtained crude
intermediate in 125 mL of ether was added by portions
and cooling in ice 4.3 g (0.11 mol) of LAH. The mixture
was stirred for 2 h. After dropwise addition of 20 mL of
saturated Na2SO4 solution the mixture was stirred for 2 h
and filtered. The filtrate was extracted with dilute aqueous
HCl and the acid layer was alkalized with NaOH and
extracted with ether. The extract was dried over anhy-
drous Na2SO4 and the ether was evaporated. The residual
oil was dissolved in 20 mL of ethanol and added to a
solution of 7.5 g (0.06 mol) of oxalic acid dihydrate in
20 mL of ethanol. Ether was added and the solution was
cooled over night. The yielded oxalate salt was filtered off
and dried; yield 3.0 g (30%); GC purity 98%, MS (CI):
m/e (M+ + 1) 201. Calc. for C10H14F2N2 200.24. Recrys-
tallization from ethanol–ether gave an analytical pure
5.1.3. 2,6-Difluoro-4-(N-methylpivalolylamino)benz-
aldehyde 11
To a stirred solution of 4.54 g (0.02 mol) of compound
10 in 40 mL of dry THF was added under nitrogen, while
cooling in a solid CO2–acetone bath, 12.5 mL of a 1.6 M
1
sample, m.p. 208–209 °C; H NMR (D2O) δ 6.3 (d, 2H,
JF,H = 9 Hz, Ar), 3.4–3.6 (m, 1H, CH), 2.8 (d, 2H, JH,H
=

146
7 Hz, CH2), 2.6 (s, 3H, CH3), 1.2 (d, 3H, JH,H = 7 Hz,
α-CH3). Anal. (C10H14F2N2•0.5(COOH)2) C, H, N.
cooling in a solid CO2–acetone bath 18.0 mL of a 1.6 M
n-BuLi hexane solution. After 1.5 h 4.0 mL (0.05 mol) of
dry DMF was added. The mixture was stirred under
cooling for 1.5 h and then poured into dilute H2SO4 and
extracted with ether. The extract was washed with water,
dried and the solvent was removed under reduced pres-
sure. The residual semi-solid product was recrystallized
from ethanol to yield 5.5 g (82%) of the desired aldehyde,
m.p. 105–106 °C; 1H NMR (CDCl3) δ 10.2 (s, 1H,
CHO), 7.4–7.1 (m, 10H, 2C6H5), 6.2 (d, 2H, JF,H = 12.9
Hz, Ar), 4.7 (s, 4H, 2CH2); Anal. (C21H17F2NO) C, H, N.
5.1.5. 3,5-Difluoro-N-pivaloylaniline
To a stirred solution of 5.0 g (0.039 mol) 3,5-
difluoroaniline in 30 mL of dry pyridine was added
5.0 mL (0.04 mol) of pivaloyl chloride in one portion.
After 2 h at room temperature the mixture was poured
into 500 mL of water. The precipitate was filtered off and
dried: yield 8.2 g (98%); m.p. 110–111 °C (m.p.
113.5 °C [15]) MS (CI): m/e (M+ + 1) 214.
5.1.6. 2,4-Difluoro-6-(pivaloylamino)benzaldehyde 14
To a stirred solution of 4.26 g (0.2 mol) of 3,5-
difluoro-N-pivaloylaniline in 40 mL of dry THF was
5.1.9. 3,5-Difluoro-4-(2-nitropropen-l-yl)-N,N-diben-
zylaniline 17
A mixture of 5.3 g (0.016 mol) of compound 16,
2.0 mL (0.028 mol) of nitroethane and 1.5 g of NH4OAc
was heated under reflux while stirring for 5 h. After
cooling the obtained red precipitate was filtered off and
recrystallized from ethanol to yield 4.2 g (67%) of 17,
added under nitrogen while cooling in
a solid
CO2–acetone bath, 25 mL of a 1.6 M solution of n-BuLi
in hexane. After 1 h at –80 °C, 4.0 mL of dry DMF
(0.05 mol) in 10 mL of dry THF was added at the same
temperature and the mixture was stirred for 1 h. The
mixture was poured into dilute hydrochloric acid and ice
and extracted with ether and the solvent was evaporated.
To the residue was added 5.0 g of NaHSO4 and 75 mL of
water. The mixture was stirred overnight at room tem-
perature and washed with ether. The water layer was
separated and 10 g of K2CO3 was added. The mixture
was stirred and cooled in an ice-bath. The formed
precipitate was filtered off and yielded, after recrystalli-
zation three times from dilute ethanol, 1.0 g (21%) of the
pure aldehyde, m.p. 67–68 °C. MS (CI): m/z (M+ + 1)
1
m.p. 96–97 °C. MS (CI); m/z (M+ + 1) 395; H NMR
(CDCl3) δ 7.9 (s, 1H, CH), 7.4–7.2 (m, 10H, 2C6H5), 6.3
(d, 2H, JH,F = 12.0 Hz, Ar), 4.7 (s, 4H, 2CH2), 2.2 (s, 3H,
CH3). Anal. (C23H20F2N2O2) H, N; C: calc. 70.04; found
69.5 69.4.
5.1.10. 1-(2,6-Difluoro-4-aminophenyl)-2-aminopro-
pane sesquioxalate monohydrate 3
A solution of 13.5 g (0.034 mol) of compound 17 in
200 mL of dry ether was added dropwise under N2 to a
stirred mixture of 4.5 g (0.12 mol) of LAH in 100 mL of
dry ether. The mixture was stirred and heated under reflux
for 4 h. After dropwise addition of 20 mL of saturated
Na2SO4 solution while stirring and cooling in ice, the
mixture was filtered. The filtrate was collected, and the
solvent was evaporated under reduced pressure. The
residual oil was dissolved in mixture of 25 mL of 12 M
HCl and 140 mL of 50% AcOH and was hydrogenated
with H2/Pd-C (10%) at normal pressure and at approxi-
mately 50 °C. When the uptake of H2 had stopped, the
catalyst was filtered off and the filtrate was evaporated to
about 50 mL and alkalized with NaOH. The mixture was
extracted with ether and the extract was dried over
Na2SO4 and the solvent was removed. The residual oil
was dissolved in 100 mL of ethanol and added to a
solution of 8.5 g (0.068 mol) of oxalic acid dihydrate in
100 mL of ethanol. The formed precipitation was dis-
solved by dropwise addition of water while stirring at
reflux temperature. After cooling the yielded oxalate was
filtered off and again recrystallized from dilute ethanol:
yield 4.3 g (37%); m.p. > 180 °C (dec.). MS (CI): m/z
1
242; H NMR (CDCl3) δ 11.8 (s, 1H, NH), 10.3 (s, 1H,
CHO), 8.4 (dt, 1H, Ar), 6.6 (m, 1H, Ar), 1.2 (s, 9H,
3CH3); Anal. (C12H13F2NO2) C, H, N.
5.1.7. N,N-Dibenzyl-3,5-difluoroaniline 15
To
a mixture of 10.0 g (0.078 mol) of 3.5-
difluoroaniline, 22.0 g (0.16 mol) of K2CO3 and a cata-
lytical amount of KI in 30 mL of DMF was added
dropwise while stirring and heating at 100 °C, 20.7 mL
(0.18 mol) of benzyl chloride. The mixture was heated
overnight at 100 °C and 5 g K2CO3 and 5.0 mL of benzyl
chloride were added. The mixture was stirred and heated
for additional 6 h at 100 °C and then poured into 500 mL
of an ice-water mixture. The obtained solid was filtered
and recrystallized from dilute ethanol: yield 19.3 g
1
(80%), m.p. 74–75 °C. MS (CI) m/z (M+ + 1) 310; H
NMR (CDCl3) δ 7.4–7.2 (m, 10H, 2C6H5), 6.3–6.1 (m,
3H, Ar), 4.6 (s, 4H, 2CH2); Anal. (C20H17F2N) C, H, N.
5.1.8. 2,6-Difluoro-4-(N,N-dibenzylamino) benzalde-
hyde 16
To a stirred solution of 6.2 g (0.02 mol) of compound
15 in 40 mL of dry THF under N2 was added while
1
(M+ + 1) 187, H NMR (CDCl3) δ 6.2 (d, 2H, JH,F = 10
Hz, Ar), 3.1–3.2 (m, H, CH), 2.6 (dd, 1H, Jgem = 13 Hz,

147
JH,H = 6 Hz, CH2), 2.5 (dd, 1H, Jgem = 13 Hz, JH,H = 8
Hz, CH2), 1.1 (d, 3H, JH,H (6 Hz, CH3). Anal.
(C9H12F2N2•1.5(COOH)2 H2O) C, H, N. Anal. after
drying at 140 °C (C19H12F2N2•1.5(COOH)2) C, H, N, F.
5.1.12. 2-Chloro-N-[1-(4-N-methyl-N-benzylamino-2-
fluorophenyl)-2-propyl]acetamide 19
A solution of 27.6 g (0.092 mol) of compound 18 in
200 mL of dry THF was added dropwise while stirring
under N2 to 12.5 g of LAH (0.33 mol) in 200 mL of dry
ether. The mixture was stirred and heated under reflux for
2 h. After dropwise addition of 60 mL of saturated
Na2SO4 solution while stirring and cooling in ice, the
mixture was filtered. The filtrate was dried over Na2SO4
and the ether was evaporated. The residual oil (24.8 g)
was dissolved in 200 mL of toluene and a solution of
16.0 g NaOH (0.4 mol) in 100 mL of water was added. To
the mixture was added dropwise while stirring and
cooling in ice-water, 20 mL (0.24 mol) of chloroacetyl
chloride. The mixture was stirred at 0 °C for 1 h and the
toluene layer was separated, dried and the solvent was
evaporated. The residual oil solified gradually upon
standing. GC/MS analysis of the obtained product
(24.1 g, 66%) showed the presence of 88% of the target
compound 19. Recrystallization from ethanol–isopropyl
ether yielded 16.6 g (52%) of the pure amide, m.p.
85–86 °C; MS (EI): m/z 348. Anal. (C19H22FClN2O) C,
H, N.
5.1.11. 3-Fluoro-4-(2-nitropropen-l-yl)-N-methyl-N-
benzylaniline 18
A solution of 26.0 g (0.23 mol) of 3-fluoro-aniline in
70 mL of 98% formic acid was refluxed for 2h and then
poured into 500 mL of water. The obtained
3-fluoroformanilide was filtered off, washed with water
and dried: yield 23.0 g (72%); m.p. 61–62 °C (m.p.
63–64 °C) [16]. The product (23.0 g, 0.16 mol) was dis-
solved in 100 mL of dry THF and the solution was added
dropwise under N2 while stirring to 15.2 g of LAH
(0.4 mol) in 200 mL of ether. The mixture was stirred and
heated under reflux for 4 h. After dropwise addition of
70 mL of saturated Na2SO4 solution while stirring and
cooling in ice, the mixture was filtered. The filtrate was
dried over Na2SO4 and the ether was removed. The
residual, crude 3-fluoro-N-methylaniline was dissolved in
50 mL of DMF and 44.0 g of K2CO3 (0.32 mol) and a
catalytical amount of KI was added. While stirring and
heating 20.7 mL (0.18 mol) of benzyl chloride was
added. The mixture was stirred overnight at 100 °C and
then poured into 500 mL of water and extracted with
ether. The extract was washed with water, dried with
Na2SO4 and the ether was evaporated. The residual oil
(32.0 g) was analyzed by means of GC/MS and was
shown to contain 94% of 3-fluoro-N-methyl-N-
benzylaniline. The crude product (32.0 g, 0.14 mol) was
dissolved in 52.0 mL of DMF and 15 mL of POCl3 was
added dropwise while stirring and cooling with tap water.
The mixture was heated for 2 h at 100 °C and poured into
500 mL of an ice-water mixture. The mixture was made
basic with 10 M NaOH and extracted with ether. The
extract was washed with water, dried and the ether was
evaporated. GC/MS analysis of the residue (36.0 g of an
oil) showed the presence of 81% of the intermediate,
4-(N-benzyl-N-methylamino)-2-fluorobenzaldehyde. The
crude product (36.0 g, 0.12 mol) was dissolved in 200 mL
of ethanol and 7.0 g of ammonium actate and 10.0 mL of
nitroethane were added. The mixture was heated under
reflux for 5 h. After cooling overnight, the obtained
precipitate was filtered off and washed with ethanol
yielding 27.8 g (77%) of a red compound, containing
96% (GC/MS) of the wanted nitro compound. Recrystal-
lization from ethanol gave an analytical pure sample,
m.p. 91–92 °C; MS (EI): m/z (M+) 300. Anal.
(C17H17FN2O2) C, H, N.
5.1.13.
N’-[1-(4-N-benzyl-N-methylamino-2-fluoro-
phenyl)-2-propyl]-N,N-dibenzylglycinamide 20
A mixture of 16.6 g (0.047 mol) of compound 19, a
catalytic amount of KI, 9.6 mL (0.05 mol) of dibenzy-
lamine and 6.9 g (0.05 mol) of K2CO3 in 25 mL of DMF
was stirred for 4 h at 100 °C. To the mixture was added
1 L of water and the separated oil was extracted with
ether. The ether extract was dried with Na2SO4 and the
solvent was evaporated. The obtained oil solified gradu-
ally and showed, when analyzed, the presence of 95% of
compound 20: yield 22.9 g (96%), m.p. 85–87 °C. Re-
crystallization from isopropyl ether–petroleum ether gave
an analytical pure sample, m.p. 87–88 °C; MS (CI: m/z
(M+ + 1) 510. Anal. (C33H36FN3O) C, H, N.
5.1.14.
N’-[1-(4-Methylamino-2-fluorophenyl)-2-
propyl]glycinamide acetate 7
Pd/C (10%) was added to a solution of 10.2 g
(0.02 mol) of compound 20 in 100 mL of acetic acid and
hydrogenated at 50 °C and normal pressure until no more
hydrogen was absorbed (3 h). The catalyst was filtered off
and the solvent was evaporated under reduced pressure.
The residue was dissolved in 20 mL of hot ethanol and
ether was added until a weak turbidity was observed. The
mixture was cooled overnight and the yielded precipita-
tion was filtered off and washed with ether: yield 3.3 g

148
1
(55%), m.p. 112–113 °C. H NMR (D2O) δ 7.0–7.1 (m,
1H, Ar), 6.5–6.6 (m, 2H, Ar), 4.1–4.2 (m, 1H, CH), 3.7
(d, 1H, Jgem = 16 Hz, COCH2), 3.6 (d, 1H, Jgem = 16 Hz,
COCH2), 2.8 (dd, 1H, Jgem = 14 Hz, JH,H = 5.5 Hz, CH2),
2.6 (dd, 1H, Jgem = 14 Hz, JH,H = 8.2 Hz, CH2), 2.7 (s,
3H, N–CH3), 1.9 (s, 3H, HAc), 1.15 (d, 3H, JH,H = 6.6
Hz, α-CH3); MS (EI): m/z (M+) 239. Anal.
(C12H18FN3O•CH3COOH) H; C: calc. 56.17; found 55.5,
55.2; N: calc. 14.04, found 13.5, 13.5.
5.1.16. N’-[1-(4-N-Benzyl-N-methylamino-2,6-difluo-
rophenyl)-2-propyl]-N,N-dibenzylglycinamide 13
To a solution of 22.7 g (0.087 mol) of compound 12 in
300 mL of ethanol was added 6.0 g of ammonium acetate
and 8.6 mL of nitroethane. The mixture was heated under
reflux for 5 h and the ethanol was removed at reduced
pressure. After the addition of 500 mL of water the
mixture was extracted with ether. The extract was washed
with water, dried with Na2SO4 and the ether was evapo-
rated. The GC/MS analysis of the residual red oil (27.1 g)
showed that the product contained 92% of the interme-
diate nitrostyrene.
5.1.15. 4-(N-Benzyl-N-methylamino)-2,6-difluorobenz-
aldehyde 12
A solution of 27.1 g (0.08 mol) of the yielded oil in
150 mL of dry THF was added dropwise under N2 while
stirring to 12.0 g (0.3 mol) LAH in 200 mL of dry ether.
The mixture was stirred and heated under reflux for 2 h.
After dropwise addition of 60 mL of saturated Na2SO4
solution while stirring and cooling in ice, the mixture was
filtered. The residual oil (24.2 g) was shown (GC/MS) to
contain 89% of the desired amine. The product (24.1 g,
0.07 mol) was dissolved in 200 mL of toluene and a
solution of 14.0 g (0.36 mol) of NaOH in 100 mL of
water was added. To the mixture was added by portions
while stirring and cooling in ice, 17.7 mL (0.22 mol) of
chloroacetyl chloride. The mixture was stirred for 1 h and
the toluene layer was separated, dried and the solvent was
evaporated. The residual oil was stirred with isopropyl
ether and the obtained solid was filtered. GC/MS analysis
of the yielded product (19.6 g) showed the presence of
92% the intermediate amide. The compound (19.6 g, 0.05
mole) was dissolved in 25.0 mL of DMF and a catalytical
amount of KI, 10.0 mL (0.053 mol) of dibenzylamine and
7.0 g of K2CO3 were added. The mixture was stirred for
5 h at 100 °C and then added to 1 L of water and
extracted with ether. The extract was washed with water,
dried and the solvent was evaporated to yield 28.0 g of an
oil containing 92% (GC) of the intermediate dibenzy-
lamine. An analytical sample, obtained by recrystalliza-
tion of the crude substance from isopropyl ether–petro-
leum ether melted at 87–88 °C: MS (EI): m/z (M+) 528;
Anal. (C33H35F2N3O) C, H, N.
A solution of 18.7 g (0.12 mol) of compound 9 in
100 mL of THF was added dropwise under N2 while
stirring to 11.4 g of LAH (0.3 mol) in 150 mL of ether.
The mixture was stirred under reflux for 4 h. After
dropwise addition of 60 mL of saturated Na2SO4 solution
the mixture was filtered, dried with Na2SO4 and the ether
was evaporated. The residual oil (17.0 g) was analysed by
means of GC/MS and was shown to contain 98% of
2,6-difluoro-N-methylaniline. The product (17.0 g,
0.12 mol) was dissolved in 50 mL of DMF and 33.2 g
(0.24 mol) K2CO3 and a catalytical amount of KI was
added. While stirring and heating at 100 °C, 32.2 mL
(0.28 mol) of benzyl chloride was added dropwise. The
mixture was stirred overnight at 100 °C and 200 mL of
CHCl3 was added. After filtration the solvent was evapo-
rated. The residual oil was dissolved in n-hexane and the
solution was acidified with hydrochloric acid in ether.
The precipitated hydrochloric salt was filtered off,
washed with ether, suspended in 500 mL of water which
was made alkaline with NaOH. The obtained oil was
extracted with ether, dried with Na2SO4 and the ether was
evaporated. The residue (23.6 g) was shown by GC/MS
to contain 95% of 2,6-difluoro-N-benzyl-N-methyl-
aniline. The oil solified gradually, m.p. 46–48 °C. To a
stirred solution of 23.2 g (0.1 mol) of the obtained prod-
uct in 200 mL of dry THF was added under nitrogen,
while cooling in a solid CO2–acetone bath, 90 mL of a
1.6 M solution of n-BuLi in hexane. After 1.5 h at –80 °C,
20 mL (0.15 mol) of dry DMF was added at the same
temperature and the mixture was stirred for 1.5 h. The
solution was poured into dilute H2SO4 and ice and the
obtained aldehyde was filtered off and washed with water.
The product, yielded after recrystallization from dilute
ethanol, 22.9 g (88%) of the pure aldehyde, m.p.
5.1.17. N’-[1-(4-Methylamino-2,6-difluorophenyl)-2-
propyl]glycinamide acetate 5
This compound was prepared analogously to com-
pound 7 from 27.5 g (0.048 mol) of the crude amide 13:
yield 6.6 g (43%); m.p. 106–107 °C: MS (EI): m/z (M+)
257. 1H NMR (D2O) δ 6.35 (d, 2H, JH,F = 10.2 Hz, Ar),
4.1–4.2 (m, 1H, CH), 3.7 (d, 1H, Jgem = 16 Hz, COCH2),
3.6 (d, 1H, Jgem = 16 Hz, COCH2), 2.8 (dd, 1H, J =
unres., CH2), 2.65 (dd, 1H, J = unres., CH2), 2.7 (s, 3H,
83–84 °C: MS (EI): m/z (M+) 261; H NMR (CDCl3) δ
10.1 (s, 1H, CHO), 7.1–7.4 (m, 5H, Ar), 6.2 (d, 2H, JH,F
= 13 Hz, Ar), 4.6 (s, 2H, CH2), 3.1 (s, 3H, CH3). Anal.
(C15H13F2NO) C, H, N.
1

149
NCH3), 1.8 (s, 3H, AcOH), 1.15 (d, 3H, JH,H = 6.7 Hz,
α-CH3). Anal. (C12H17F2N3O•CH3COOH) C, H, N.
turbidity was observed. The mixture was cooled over-
night and the yielded precipitate was filtered off and
washed with ether. The crude acetate salt (6.7 g), m.p.
99–102 °C, was recrystallized from THF to yield 5.0 g
(35%) of the pure compound, m.p. 102–103 °C: MS (EI):
5.1.18.
N’-[1-(4-Dimethylamino-2-fluorophenyl)-2-
propyl]glycinamide oxalate 8
1
m/z (M+) 225. H NMR (D2O) δ 6.9–6.8 (m, 1H, Ar),
To a mixture of 1.2 g (0.004 mol) of 1-(4-dime-
thylamino-2-fluorophenyl)-2-aminopropane dihydrochlo-
ride, 25.0 mL of water, 50.0 mL of ether and 2.0 g NaOH
(0.05 mol) was added dropwise while stirring and cooling
in ice-water, 3.5 mL (0.04 mol) of chloroacetyl chloride.
The mixture was stirred at 0 °C for 0.5 h and the ether
layer was separated, dried and the ether was evaporated.
The residual oil was dissolved in 50.0 mL of ethanol,
50.0 mL of a 25% aqueous NH3 and left for 5 days at
room temperature. The ethanol was evaporated and the
residue was extracted with ether. The ether layer was
dried and the solvent was evaporated. The residual oil
was dissolved in 50.0 mL of hot ethanol and a solution of
1.26 g (0.01 mol) of oxalic acid dihydrate was added. The
mixture was cooled overnight and the precipitate was
filtered off and recrystallized from ethanol to give 8
(0.5 g, 37%): m.p. 183–184 °C. 1H NMR (D2O) δ
7.1–7.2 (m, 1H, Ar), 6.8–7.0 (m, 2H, Ar), 4.0–4.1 (m, 1H,
CH), 3.6 (d, 1H, Jgem = 16.1 Hz, COCH2), 3.5 (d, 1H,
Jgem = 16.1 Hz, COCH2), 2.9 (s, 6H, (NCCH3)2), 2.8 (d,
1H, J = unres., CH2), 2.6 (d, 1H, J = unres., CH2), 1.0 (d,
3H, JH,H = 6.8 Hz, α-CH3); MS (EI): m/z (M+) 253. Anal.
(C13H20FN3O•(COOH)2) C, H, N.
6.4–6.3 (m, 2H, Ar), 4.0–3.9 (m, 1H, CH), 3.5 (d, 1H,
Jgem = 15.8 Hz, COCH2), 3.4 (d, 1H, Jgem = 15.8 Hz,
COCH2), 2.6 (dd, 1H, Jgem = 14 Hz, JH,H = 5.3 Hz, CH2),
2.4 (dd, 1H, Jgem = 14 Hz, JH,H = 8 Hz, CH2), 1.2 (d, 3H,
J
H,H = 6.7 Hz, CH3). Anal. (C11H16FN3O•CH3COOH) C,
H, N.
5.1.20. 1-(2-Fluoro-4-methylaminophenyl)-2-amino-
propane acetate 4
Compound 18 (45.9 g, 0.15 mol) was reduced with
21.0 g LAH as described for compound 17 13.6 g
(0.05 mol) of the obtained crude oil (42.6 g, purity 88%)
was dissolved in 100 mLAcOH and hydrogenated similar
to compound 17. The yielded oil was dissolved in 50 mL
of ethanol and ether was added until a weak turbidity was
observed. After cooling the acetate salt (8.5 g, 70%) was
filtered off, m.p. 148–149 °C: MS (EI): m/z (M+) 182. 1H
NMR (D2O) δ 7.05–7.15 (m, H, Ar), 6.5–6.6 (m, 2H, Ar),
3.5–3.6 (m, H, CH), 2.85 (d, 2H, JH,H = 7.1 Hz, CH2), 2.7
(s, 3H, NCH3), 1.9 (s, 3H, AcOH) 1.3 (d, 3H, JH,H = 6.7
Hz, α-CH3). Anal. (C10H15FN2•CH3COOH) C, H, N.
5.1.21. N’-(S)-[1-(4-Dimethylamino-2-methylphenyl)-
2-propyl]-(S)-valinamide dihydrochloride 21
5.1.19.
N’-[1-(4-Amino-2-fluorophenyl)-2-propyl]
glycinamide acetate 6
To a stirred solution of 5.0 g (0.02 mol) of
N-carbobenzyloxy-(S)-valine in 25.0 mL of dry THF was
added 3.2 g (0.02 mol) of CDI. The reaction mixture was
heated at reflux temperature until evolution of CO2
ceased and 3.8 g (0.02 mol) of (S)-1a was added. The
mixture was stirred at reflux temperature over night. The
THF was evaporated and 200 mL of water was added.
The obtained precipitate was filtered off and washed with
water. Recrystallization from aqueous ethanol yielded
5.3 g of pure N-carbobenzyloxy-N’-(S)-[1-(4-dime-
thylamino-2-methylphenyl)-2-propyl]-(S)-valinamide;
MS (EI): m/z (M+) 425.
To a solution of 30.0 g (0.08 mol) of 1-(4-diben-
zylamino-2-fluorophenyl)-2-aminopropane in 100.0 mL
of toluene was added a solution of 13.2 g (0.33 mol) of
NaOH in 100 mL of water. To the mixture was added
dropwise while stirring and cooling in ice-water, 17.0 mL
(0.2 mol.) of chloroacetyl chloride. The mixture was
stirred at 0 °C for 1 h and the obtained precipitate was
filtered off, washed with water and dried. The obtained
product (21.3 g) was dissolved in 25 mL of DMF and
10.0 mL (0.053 mol) of dibenzylamine, a catalytical
amount of KI and 7.0 g (0.05 mol) of K2CO3 were added.
The mixture was stirred for 4 h at 100 °C and was then
diltuted to 500 mL with water and extracted with ether.
The extract was washed with water, dried with Na2SO4
and the solvent was evaporated. The residual oil (29.8 g)
was dissolved in 100 mL of AcOH and was hydrogenated
with H2/Pd-C 10% at normal pressure and approximately
50 °C. When the uptake of hydrogen had stopped, the
catalyst was filtered off and the acetic acid was evapo-
rated at reduced pressure. The residue was dissolved in
20 mL of hot ethanol and ether was added until a weak
The obtained product (5.3 g) was dissolved in 100 mL
of acetic acid and hydrogenated with H2/Pd-C 10% at
normal pressure and approximately 50 °C. The catalyst
was filtered off and the acetic acid was evaporated at
reduced pressure. The residue was dissolved in hot
ethanol and acidified by the addition of hydrogen chloride
in ether. To the solution was added petroleum ether (b.p.
40–60 °C) and the mixture was cooled over night. The
yielded precipitation was filtered off and washed with
1
ether: yield 2.8 g (62%), m.p. 246–248 °C: H NMR

150
(D2O) δ 7.46 (d, 1H, J = unres., Ar), 7.45 (d, 1H, JH,H
8.1 Hz, Ar), 7.40 (d, 1H, JH,H = 8.1 Hz, Ar), 4.43 (m, 1H,
CH), 3.67 (d, 1H, JH,H = 5.4 Hz, CH), 3.38 (s, 3H, CH3),
=
5.2.2. Tissue preparation in in vitro experiments
The rats were killed by decapitation and the whole
brain, except the cerebellum was dissected out on ice.
The mitochondria were prepared as described by Ask et
al. [16].
3.28 (s, 3H, CH3), 3.06 (dd, 1H, Jgem = 14.4 Hz, JH,H
=
4.7 Hz, CH2), 2.85 (dd, 1H, Jgem = 14.4 Hz, JH,H = 10.7
Hz, CH2), 2.47 (s, 3H, CH3), 1.96 (m, 1H, CH), 1.33 (d,
3H, JH,H = 6.6 Hz, CH3), 0.73 (d, 3H, JH,H = 6.9 Hz,
CH3), 0.64 (d, 3H, JH,H = 6.9 Hz, CH3); [α]20 = + 23.2 (c
= 1, H2O). Anal. (C17H29N3O•2HCl) C, H, Cl, N; N: calc.
11.53; found 10.9, 10.8, 10.9.
5.2.3. Tissue preparation in ex vivo experiments
The rats were killed by decapitation 48 h after the
administration of the test compounds when the effect of
these were presumed to have disappeared. The hypo-
thalami and striata were rapidly dissected out, weighed,
and homogenized in 20 volumes of ice-chilled 0.25 M
sucrose in small all-glass homogenizers. The homoge-
nates were centrifuged at 800 g for 10 min and the
supernatants were used in the MAO assay.
5.1.22. N’-(S)-[1-(4-Dimethylamino-2-methylphenyl)-
2-propyl]-(R)-valinamide dihydrobromide 22
The intermediate compound N-carbobenzyloxy-N’-(S)-
[1-(4-dimethylamino-2-methylphenyl)-2-propyl]-(R)-
valinamide was prepared as described for compound 21.
From 3.8 g (0.02 mol) of amiflamine and 5.0 g (0.02 mol)
of carbobenzyloxy protected (R)-valine was thus obtained
6.2 g amide melting at 175–77 °C. The yielded com-
pound was added by portions to 20 mL of 30% HBr in
acetic acid. The mixture was stirred at room temperature
for 0.5 h. Ether was added and the obtained hydrobromic
salt was filtered off and recrystallized twice from etha-
nol–petroleum ether: yield 4.6 g (68%), m.p.
5.2.4. MAO activity in vitro
The incubation medium consisted of 50 mL of mito-
chondrial suspension, 100 mL of the test compound or
distilled water, 820 mL of 0.11 M sodium phosphate
buffer pH 7.4 and after 10 min preincubation 25 mL of
the substrate, either [14C]5-HT (50 µM final concentra-
tion) or [14C]PEA (2.5 µM final concentration) [17]. The
incubation was continued for 5 min, the reaction was
stopped by the addition of 1 mL of 1 M HCl. The acid
metabolites were extracted into 6 mL of ethyl acetate by
vigorous shaking in a multitube vortex mixer (Model
2601, Scientific Manufactoring Industries). After cen-
trifugation 4 mL of ethyl acetate was transferred to a
counting vial containing 1 mL of ethanol and 10 mL of
liquid scintillation cocktail (Ultima Gold, Packard). The
radioactivity was measured in a liquid scintillation
counter (LS 6000TA, Beckman).
1
225–226 °C: MS (EI): m/z (M+) 291. H NMR (D2O) δ
7.49 (s, 1H, Ar), 7.44 (br.m., 2H, Ar), 4.26 (m, 1H, CH),
3.76 (d, 1H, JH,H = 5.8 Hz, CH), 3.32 (s, 6H, 2CH3), 2.98
(dd, 1H, Jgem = 14.1 Hz, JH,H = 7.8 Hz, CH2), 2.94 (dd,
1H, Jgem = 14.1 Hz, JH,H = 6.9 Hz, CH2), 2.48 (s, 3H,
CH3), 2.25 (m, 1H, CH), 1.26 (d, 3H, JH,H = 6.6 Hz,
CH3), 1.06 (d, 6H, JH,H = 6.9 Hz, 2CH3); [α]20 = –4.2 (c
= 1, H2O). Anal. (C17H29N3O•2HBr) C, H, Br, N.
5.2.5. MAO activity ex vivo
The deamination of [14C]5-HT, [3H]NA and [14C]DA
by the synaptosomal preparation was determined as
described previously by [7]. After a 10-min preincubation
of 50 mL of the synaptosome-rich supernatants in
925 mL of Krebs–Henseleit’s buffer, pH 7.4, containing
5.6 mM glucose, 1.1 mM ascorbic acid, and 0.13 mM
disodium edetate, the inkubation was continued for a
further 10 min at 37 °C with [14C]5-HT (0.1 µM) or
[14C]DA (0.25 µM) in the absence and presence of 0.12
µM citalopram (5-HT) or 0.1 µM GBR 12909 (DA). In
the [3H]NA (0.25 µM) experiments 200 mL of the hypo-
thalamic supernatants was used in the absence and
presence of maprotiline (3 µM). The deaminated products
were extracted into ethyl acetate in accordance with the in
vitro experiments, and the monoamine oxidase activities
were calculated from the radioactivities.
5.2. Pharmacology
5.2.1. Animals
Male Sprague–Dawley rats (Alab-SD) weighing
160–200 g were obtained from B& K International
(former Alab Laboratorietjänst AB), Sollentuna, Sweden.
They were housed in plastic cages with sawdust in groups
of 5 under constant temperature and lighting (6 a.m.–6
p.m.) and were allowed free access of food and water. In
the in vivo experiments the animals were deprived of
food on the night before the oral administration of the test
compound. Phenelzine sulphate 4 mg/kg s.c. in the ex-
periment with serotonin (5-HT) and noradrenaline (NA)
and 2.5 mg/kg s.c. in the experiment with dopamine (DA)
was injected 1 h after the test compound. The rats were
then resupplied with food.
The MAO inhibition inside and outside the aminergic
neurons was estimated from the protection against the

151
irreversible action of phenelzine as described by Green
and El Hait [13]. The percent inhibition produced by the
reversible inhibitors was calculated according to the
following formula:
References
[1] Johnston J.P., Biochem. Pharmacol. 17 (1968) 1285–1297.
[2] Yang H.Y.T., Neff N.H., J. Pharmacol. Exp. Ther. 187 (1973)
365–371.
[1 – ln(100/t) / ln(100/p)] × 100
[3] Cesura A.M., Pletscher A., In: Junker E. (Ed.), Progress in Drug
Research, Vol. 38, Birkhauser Verlag, Basel, 1992, pp. 171–297.
in which t is the MAO activity in the synaptosomes from
the animals treated with the test compound + phenelzine,
and p is that from the rats treated with saline +
phenelzine, the activity expressed in percentage of that in
the control animals. The ED50 values were estimated
from log–response curves based on at least three doses
with four rats in each dose group.
[4] Florvall L., Ask A.L., Ögren S.O., Ross S.B., J. Med. Chem. 21
(1978) 56–63.
[5] Ask A.L., Fagervall I., Florvall L., Ross S.B., Ytterborn Y., Br. J.
Pharmacol. 85 (1985) 683–690.
[6] Ask A.L., Fagervall I., Florvall L., Ross S.B., Ytterborn S.,
Neuropharmacology 25 (1985) 33–40.
[7] Ask A.L., Fagervall I., Ross S.B., Naunyn-Schmiedeberg’s Arch.
Pharmacol. 324 (1983) 79–87.
5.2.6. Determination of 5-HT, NA, DA and metabolites
The concentrations of 5-HT, 5-HIAA, NA, DA,
DOPAC and HVA in hypothalamus, hippocampus and
striatum were determined by use of high performance
liquid chromatography with electrochemical detection
according to the method of Magnusson et al. [18] The
mobile phase was 0.1 M phosphate buffer (pH 2.5):
methanol: acetonitrile (89:9:2 v/v) containing 1 mM octyl
sulphate. The frozen samples were weighed and homog-
enized in 0.1 M perchloric acid, containing 2.5 mM
sodium bisulphite, 1 mM ethylendiaminetetraacetic acid
(EDTA) and isoprenaline as internal standard. The super-
natants were injected directly onto a Supelcosil C 18
(3 µM) column, connected to a detector (ESA Coulochem
5100), set to 0.05/0.30 V.
[8] Florvall L., Fagervall I., Ask A.L., Ross S.B., J. Med. Chem. 29
(1986) 2250–2256.
[9] Fuhrer W., Gschwend H.W., J. Org. Chem. 44 (1979) 1133–1136.
[10] Florvall L., Persson M.L., Acta Chem. Scand. B38 (1982)
141–146.
[11] Grassmann W., Wunsch E., Ber. 91 (1958) 462–465.
[12] Adkins H., Ross R.M., Schroeder D.C., Mahoney C.L., Gilbert
W.W., J. Am. Chem. Soc. 6 (1954) 147–152.
[13] Green A.L., ElHait A.S., Biochem. Pharmacol. 29 (1980)
2781–2789.
[14] McTavish D., Goa K.L., Drugs 38 (1989) 757–777.
[15] Thornton T.J., Jarman M., Synthesis (1990) 295–299.
[16] Pettit G.R., Kalnins M.V., Liu T.M.H., Thomas E.G., Parent K., J
Org. Chem. 26 (1961) 2563–2566.
Acknowledgements
[17] Ask A.L., Hellström W., Norrman S., Ögren S.O., Ross S.B.,
Neuropharmacology 21 (1982) 299–308.
We thank Anders Andersson, Astra Arcus AB, for
valuable help with the NMR spectra.
[18] Magnusson O., Nilsson L.B., Westerlund D., J. Chromatogr. 221
(1980) 237–247.